Strength in diversity? Fiscal federalism among the fifty US states

The gap between the degrees of decentralization of expenditures and revenues is even smaller for New Zealand; however, the degree of decentralization is so small in New Zealand that the gap has very little importance.

An OECD report on tax expenditures (OECD 2010) displays measures of tax expenditures as a percent of GDP in six of the countries we list in Tables 1 and 2. The figures are 7.68% for Canada, 0.64% for Germany, 1.78% for the Netherlands, 4.95% for Spain, 12.54% for the UK, and 6.47% for the USA.

In an earlier literature, Brennan and Buchanan (1980) argue that there is a causal relation between the extent of fiscal decentralization and the overall size of the public budget. They contend that competition among lower-level governments constrains the monopolistic tendencies of the public sector so that countries with highly decentralized public sectors would tend to have relatively small public sectors. This notion became known as the Leviathan hypothesis. The evidence is, however, mixed at best. Oates (1985), in the first empirical study of “Leviathan,” finds that the evidence runs counter to the Brennan–Buchanan proposition. Some later empirical work, however, suggests a more complex relationship. See, for example, Ashworth et al. (2013).

Canada is unusual in that it has both large equalizing grants and extensive subnational revenue authority. See Simeon et al. (2014).

Germany’s number is not comparable to the others because of the unique German system in which equalizing transfers occur as redistributions between states rather than grants from the central government; since Fig. 1 does not capture redistributions between states, the number for Germany does not capture the total amount of transfer revenues. See Enderlein and von Müller (2014) for a description of the differences between Germany’s form of fiscal federalism compared to others. New Zealand’s figure is very similar to the US figure, but note from Table 2 that its degree of decentralization in revenues is only 7%.

Our measure of diversity is the coefficient of variation, i.e., the standard deviation divided by the mean.

For an examination of changes in the distribution of K-12 education spending across school districts within and across states from 1972 to 1992, see Murray et al. (1998).

In early 2015, the education committee of the Nebraska legislature voted to kill a bill that would have allowed charter schools in Omaha.

Hawaii is an outlier state in many respects, not the least of which is the fact that it has only one school district.

Our description of the systems in Michigan and Ohio relies on Conlin and Thompson (2014).

Michigan school districts are constrained in terms of raising revenues for operating expenditures; however, residents can vote to raise local property taxes for capital expenditures (Conlin and Thompson 2014).

Medicaid expenditures as a share of public welfare expenditures increased from 45% in 1982 to 80% in 2012.

Rhode and Strumpf (2003), using observations on municipalities and counties in the USA from 1850 to 1990, test for evidence of Tiebout sorting by examining trends in diversity in fiscal policy. Contrary to predictions of the Tiebout model, they find that in the long run as transportation costs have fallen communities have become more alike in their fiscal policies. One possible explanation for why their results differ from ours is that we employ quite different units of analysis and analyze quite different time periods.

For interesting treatments of the evolution of the state and local sector in the USA, see Baicker et al. (2012) and Wallis (2000).

See Calsamiglia et al. (2013) for a treatment of the notion that minimal levels of spending can be achieved without losing subnational fiscal autonomy.

The Every Student Succeeds Act, signed into law in December 2015, prohibited the federal government from imposing academic standards, including the Common Core.

The Affordable Care Act illustrates the ongoing debate in the USA over the appropriate role of the central government in a fiscal federal system. Many would have preferred a federal requirement that all states expand Medicaid, but others found such a requirement overly intrusive and, in the end, states were given the choice, with 19 choosing not to expand Medicaid despite the federal government agreeing to cover virtually the entire cost of the expansion.

Recall from Table 3, the relatively low amount of diversity across regions in the three European countries analyzed.

This possibility is illustrated by Canada, where large equalizing grants exist next to extensive fiscal autonomy for the provinces.