Too Much Local Government?
Indiana Governor Mitch Daniels recently created the bipartisan Commission on Local Government Reform. In recent remarks, the Governor noted: “For its size and population, Indiana has far too much local government. Indiana has some 2,700 local units of government authorized to levy property taxes. Governing these units are more than 10,700 elected officials, 1,100 of whom assess property. Few other states have as much local government.”
Perhaps not; but there are more than 500,000 elected officials in the United States, 96 percent of whom serve in local governments. The remarks are correct insomuch as electoral density—the number of elected officials per capita or per governmental unit—varies greatly from place to place. The most electorally dense county has more than 20 times the average number of elected officials per capita.
How would we know whether Indiana or any other state has too many local governments or too few? What is the benchmark for deciding whether there too many elected officials in a jurisdiction or not enough?
In a recent paper, Christopher Berry and I suggest that more elected officials does not necessarily mean more government spending or higher taxes. Drawing on principal-agent theories of representation, we argue that electoral density presents a tradeoff between accountability and monitoring costs. Increasing the number of specialized elected offices—offices with responsibility for a single policy domain—promotes issue unbundling. For these offices, elections are a more effective way to discipline politicians, resulting in less slack between citizen preferences and government policy. But the costs of monitoring a larger number of officials may offset some of these benefits, producing greater latitude for politicians to pursue their own goals at the expense of citizen interests.
On this view, electoral density will produce a U-shaped relationship between the number of local officials and government fidelity to citizen preferences. In fact, that is what the data show. Using a county level dataset of all elected officials in the United States, we find evidence that public spending decreases as electoral density rises—up to a point—beyond which budgets grow as more officials are added within a community.