bq. This article discusses the common-pool problems that arise when multiple territorially overlapping governments share the authority to provide services and levy taxes in a common geographic area. Contrary to the traditional Tiebout model in which increasing the number of competing governments improves efficiency, I argue that increasing the number of overlapping governments results in “overfishing” from the shared tax base. I test the model empirically using data from U.S. counties and find a strong positive relationship between the number of overlapping jurisdictions and the size of the local public sector. Substantively, the “overlap effect” amounts to roughly 10% of local revenue.
That is from a recently published paper by Christopher Berry (here; here, ungated). Here is Wikipedia on the Tiebout model.
Berry notes that the Tiebout model presumes that the tax base is elastic; in short, inefficiency will either lead people to leave the local jurisdiction or reduce the value of the tax base. But the negative consequences of inefficient policy are spread among all the governments sharing the tax base, making inefficiency less costly for any particular government. Hence this conclusion:
bq. In sum, the same sorts of common-pool problems that provide the incentives for individual jurisdictions to overspend also dilute the effects of the competitive counterpressures that are thought to punish deviations from efficiency. The incentives for efficiency become weaker as more governments share the same tax base.
Another of Berry’s observations might be subtitled “Why Economics Needs Political Science”:
bq. The complex vertical layering of fiscally and politically interacting governments defies Tiebout’s institution-free model of local nonpolitics.