Local Taxation
We separate local taxation to take account of fiscal decentralization. Fiscal decentralization affects freedom in that when more taxes are raised at the local level, residents may have greater choice over their tax burden and public services. They can more easily vote with their feet—that is, they can more easily move across town lines than across state lines.
But that very ability to foot-vote varies not just by how much revenue is raised at the local level but also by the number of local jurisdictions. If local governments are spatially large, it is difficult for residents to exercise choice. When a city like Houston annexes other independent municipalities, it becomes more difficult for newcomers to the area to choose a jurisdiction to their liking. Hawaii’s single statewide school district prevents parents from moving to a district where they think the schools are better run. Because the relevant decision for a homeowner is typically over local jurisdictions within driving distance to a place of employment, the metric for variety of choice that we use is the effective number of local jurisdictions per square mile of privately owned land (we exclude publicly owned land because it is presumably not developable), in logarithms (the natural log is taken to deal with skewness and to capture diminishing marginal effects).
“Effective number of local jurisdictions” counts up the weighted sum of general-purpose local governments in each state, where the weights are the percentage of local tax revenue raised by each local government tier. For instance, if a state has 10 counties and 100 municipalities, and counties raise 40 percent of local taxes while municipalities raise 60 percent, then the state’s effective number of local jurisdictions is 10×0.4 + 100×0.6 = 64. We then divide that number by the number of square miles of private land in the state and then take the natural logarithm to reduce skew in the distribution. (This method also helps large states like Nevada and Texas relative to the New England states.) We divide by land area rather than population because we want to capture ease of moving, and residents will generally find it easier to move to a different municipality within the same labor-market area if there are more jurisdictions per square mile, even if each one is populous.
The variable for the effective number of local jurisdictions per square mile determines the weight on the local taxation variable, which therefore varies by state. It is the only variable in the index with a weight that varies by state. (The weight for local taxation reported in Figure 1 is the average for all 50 states over the 2000–2022 period.) The idea here is that high decentralization (high local taxation relative to state taxation) matters less when there are fewer jurisdictions per square mile and more when there are more. Specifically, we multiply the standard taxation weight by 0.75 for the state with the most jurisdictions per square mile (New Jersey) and give a hypothetical state with no local governments the full taxation weight and then arrange the other states linearly according to their effective number of jurisdictions per square mile. In New Jersey, we are assuming that local taxation is only three-quarters the restriction on freedom that state taxation is. In Hawaii, the most territorially centralized state, local taxation is almost the same as state taxation—the prospective homeowner has virtually no local exit option, so local taxes are only a little more likely than state taxes to reflect distinctive local preferences.
Local tax collections come from the most recent fiscal year data released by the Census Bureau (FY 2021). The numbers here represent the combined formula incorporating both the level of local taxation and the weight as determined by the number of competing local jurisdictions. As a result, the numbers in Table 1 are not directly comparable to the figures for state-level taxation already given.
The differences among states in local taxation burden have a lot to do with geography and history, not so much partisanship. Vermont looks best on local tax and worse on state tax because of its peculiar solution to centralizing school finance. Other states with low local taxation are also fiscally centralized; some of them have high state tax burdens. On the other hand, the very worst state on local tax burden—New York—also has a high state tax burden. However, other high local tax states, like Texas, Colorado, Alaska, and New Hampshire, have low state tax burdens: they are fiscally decentralized.