Ameregis


Back to American Metropolitics
Measuring a Community's Tax Capacity
In MARC's national project, tax capacity for an individual municipality is defined as the revenue that would be forthcoming if that municipality applied metropolitan-wide average tax rates to each of the tax bases actually available to it under state law. The taxes included in the analysis are property, sales and income taxes. Total revenue capacity is defined as tax capacity plus state aid. These four revenue sources represented roughly two thirds of local general revenue in the United States in 1996. The remaining third came largely from fees and charges of various sorts. Charges and fees were excluded from the analysis for two reasons, the first conceptual, the second practical. The conceptual difficulties involve defining capacity in this context. Fees and charges are assessed in many different ways. This makes it very difficult to define the base against which the fee or charge is applied (in order to compute a "tax rate"). Income might serve as a reasonable measure of the base for fees that are associated with consumption of specific services (such as water fees that are based on usage rates). However, income would not serve as a good measure of the base for impact fees -- fees that are assessed against new housing or business development to offset infrastructure or other public costs associated with the development. Some measure of the potential value of new development would serve best for those types of fees. It would be difficult in principle to measure this potential and nearly impossible in practice. The practical difficulties are primarily associated with data availability. Adequate revenue and base data for fees and charges are not generally available at the local level. In most cases, local revenue data sources do not report revenues from the different types of fees separately. When such distinctions can be made in the data, it is often because the revenues are reported separately for various enterprise funds. However, such data is often difficult to interpret because net revenues from such funds are commonly transferred from one fund to another. Reliable recent data for the potential base for such revenues is also not generally available. Local-level income data is available for all localities only in national census years. Measures of the potential base for impact or development fees are even more difficult to come by. For instance, the extent and value of land available for development is obtainable only in very select cases from land-use surveys. Comparable data for all of the municipalities in the twenty-five largest metropolitan areas is simply not available. Both conceptual and practical concerns were also involved in the decision to use metropolitan-level average tax rates (rather than national average rates) in the capacity calculation. Conceptually, one could argue that applying national average rates to local bases for all three of the taxes is preferable because that procedure would generate tax capacities that were comparable across metropolitan areas. However, this measure would provide tax capacities that were comparable only assuming that all tax bases were available to all municipalities. State laws limit local taxing powers in different ways in different states and these distinctions are important when considering local tax capacities. For instance, if Boston ranked very high among large cities in income tax capacity but relatively low in property tax capacity, it's overall tax capacity might rank fairly high. However, this would be of very limited relevance to policy-makers in Boston because under Massachusetts' laws only the property tax is available to them. In addition, it can be argued that the best universe of comparison for local tax capacities is the metropolitan area because the labor and housing markets that are most significantly affected by differences in capacities and costs (and the resulting differences in tax rates) are metropolitan in scope. However one decides the conceptual issues, practical issues mitigate against the use of national average rates applied to all three bases in all of the municipalities in the sample. In practice, tax base data is available in a state only for the taxes in common usage in that state. For instance, local sales tax base data is not generally available in states where municipalities are not granted the power to tax sales. Property tax capacity was calculated using effective tax rates (revenues as a percentage of market value) in all cases to control for the fact that assessment ratios vary across (and, in some cases, within) states. Effective sales and income tax bases were also used in order to control for differences in the coverage of the taxes across states. These decisions reflect the fact that the overall objective of the capacity calculations is to estimate each municipality's capacity to raise revenues within the context of the state laws that govern its taxing behavior. Average tax rates for each base were calculated as the mean rate for 1993 and 1998. This measure was used so that the measured growth rate in tax capacity would reflect only changes in tax bases (and not in tax rates). In the nineteen metropolitan areas, where more than one tax was available to some or all municipalities, the tax capacity calculation was adjusted to reflect the fact that revenues from one tax do not displace or augment revenues from other taxes dollar for dollar. For example, if a locality that has access to only the property tax gains access to a new tax, it is unlikely that its total tax revenues will increase by exactly the amount of the revenues from the new tax. Instead, one would expect the property tax rate to decline with the addition of the new revenues from the new tax base. Put another way, the property tax rate (and therefore revenues) will be lower in a locality that also taxes sales than it would be in the absence of the sales tax revenues. The actual difference in property tax revenues between the two situations is likely to be some fraction (between 0 and 1) of the sales tax revenues. This fraction was estimated for both sales and income taxes from a regression analysis of all of the municipalities in the 25 metropolitan areas. This analysis implied that, all else equal, local property tax revenues decline by $.373 for each $1.00 of local sales tax revenue and by $.310 for each dollar of local income tax revenues¹. These revenue displacement rates were converted into a commensurate adjustment to the property tax base, jurisdiction by jurisdiction by the following formula: PBASE^_i was then used to calculate property tax capacity. In some cases, the adjustment resulted in a PBASE^_i less than zero. In those cases, the property tax base in the community was set to zero (which implies a property tax capacity of zero). This adjustment was used in metropolitan areas where all jurisdictions used sales or income taxes and in cases where only a subset of jurisdictions used sales or income taxes but tax base data for the "extra" tax was not available. Four other situations also required special procedures: (1) areas where only one locality had access to a specific tax; (2) combined city-counties; (3) Washington D.C. (which has no state); (4) areas with classified property tax systems (different tax rates for different classes of property). (1) Localities with special access to a specific tax. In several metropolitan areas, a single municipality has the authority to levy a tax not generally available to local governments in the state or metropolitan area. This was the case in the Washington D.C., Kansas City, and St. Louis metropolitan areas. In all three cases only one city (the District of Columbia, Kansas City MO and St. Louis MO, respectively) has the authority to use a local income tax. In these cases property tax capacity in the locality with special access to the income tax was calculated using equation A.1 and income tax capacity equaled income tax revenues. (2) Combined city-counties. In this sample, the cities of Baltimore, Denver, New York, Philadelphia, San Francisco and St. Louis are combined city-counties. This means that revenues (both taxes and aid) in these cities reflect the fact that they have greater expenditure responsibilities than other municipalities in their states and metropolitan areas. To correct for this, revenue tax and aid revenues for these cities was multiplied by the statewide average local share of total local plus county revenues for the relevant source. This correction affects the tax capacity estimates only indirectly, through its impact on the calculation of the region-wide average tax rate. It affects the calculation of total revenue capacity (tax capacity plus aid) directly by decreasing the amount of aid to the relevant city that is included in the calculation. The procedure is meant to isolate revenue for municipal functions from aid for county functions. (3) Washington D.C. has no state. It therefore has expenditure responsibilities that other cities do not have and it does not share its income and sales tax bases with a state as other cities do. Sales and income tax revenues for the city were therefore scaled by the average local share of each tax for the nation as a whole. This affects the capacity calculation for Washington D.C. directly because the income tax is not in general use in the metropolitan area. This means that the tax capacity calculation for the city uses revenues from this tax directly (as described above under category 2). Washington D.C. also receives federal aid meant to replace the state aid that it does not receive. However, it is difficult to disentangle this aid from the "normal" federal aid that many cities receive (and that is not counted in total revenue capacity). In addition, it has often been argued that, for a variety of reasons, the District receives much less of this aid than it should. Federal aid to Washington D.C. is meant to compensate the city both for federal buildings, representing large amounts of potential property tax base, that are exempt from local taxation and for the aid that states normally receive. Analysts generally agree that the aid that the District receives is adequate to cover either these cost categories alone, but not both². For these reasons, federal aid was not included in the total revenue capacity calculation for Washington D.C. (4) Classified property tax systems. In states where localities may assess different tax rates to different types of property, the property tax capacity calculation was modified to allow the capacities to reflect differences in the mix of property types across places. Average tax rates were calculated separately for different types of property. Tax capacity was then calculated separately for each type of property for each jurisdiction. Total property tax capacity was the sum of capacities for each property type. ¹ The regression estimated property tax revenues per household (dependent variable) as a function of median household income, tax price (the property tax share of a median value home), intergovernmental aid per household, sales tax revenues per household, income tax revenues per household, average household size, the percentage of housing units that are owner-occupied, population density, poverty rate, age of the housing stock, and the statewide average local share of state and local direct expenditures. ² It has also been argued that Washington D.C. has been under-compensated for its welfare expenses. Welfare programs represent a major portion of federal aid to states. Historically, these funds were provided on a matching basis with the federal share determined state-by-state with low-income states receiving a larger match than high-income states. Washington D.C.'s matching rate has usually been determined independently and the rate has usually been commensurate with high-income states when actual income levels in the District would have justified a more generous match.
Back to American Metropolitics
©2002-2008 Ameregis Corporation